PREAMP technology shift? - Class A/ balanced/ dual-mono/ zero feedback vs. op-amp preamps
Hi all,
Even ten years ago, it seemed that the gold standard for solid-state pre's was some combination of the following: Class A/ balanced/ dual-mono/ zero feedback or local feedback. Recently, I'm seeing a range of well-reviewed preamps, from Schiit to Rowland, that use new, high-quality op-amps in their circuitry, to the extent that I wonder if op-amp technology, and chip manufacture in general, has gotten so good that some of the tried and true markers of quality (MOSFETs, Class A circuits) are not in the market except at the highest end, and even then, see also Jeff Rowland. Or is it that manufacturers are looking to less power hungry designs (I know that Rowland was concerned with meeting EU efficiency requirements in its latest designs) and so avoid Class A, for example? Does balanced/ dual mono seem less popular because separates are now less popular than integrateds? Have we reached a tipping point with op amps?
Here's a more focused question. I have invested in a 90s era Aragon 8008BB dual mono balanced amp, but i have an early 90s B&K PRO10-MC that I want to upgrade. Is a Schiit Freya a significant step up, or should I save my pennies for a used, more recent example of the old school: Classe CP500, Pass Labs Aleph or X2.5, or a good example of a newer approach, the Rowland Capri. Speakers are Vandersteen 2CE Sigs. Input impedance of the Aragon is 44 kOhms.
The tube route is always possible, but I like the idea of a no-fuss, very clean ss, valuing sound stage and detail over warmth, which makes me think I'd be wasting my time with the Freya (any shoutouts for its JFET stage?)
You have so many questions, I’m not entirely sure what you really want answered here. Though, it seems you are looking for a good preamp. In general op amp technology has really advanced with opamps that have come out in the last 10 years. However, usage of op amps versus discrete or Class A is entirely dependent on your application and what source you have. If you have a very laid back analog or tube source (such as a tube DAC), then a high-speed accurate bipolar preamp may be a good match. Or maybe go with a passive preamp with a higher impedance attenuator (like 100k ohms).
If you have a very solid-state DAC source that only uses op amps, then you probably want to have some sort of Class A preamp. But be aware of what you are using as some op amp sources are engineered to sound warm (like NAD or Adcom or even Emotiva XDA dacs). Many use the OPA2134 FET based op amp for this purpose (the lm49720 / lm4562 op amp is also used somewhat like this). In my opinion OPA2134 is a really poor op amp that smears the sound to make it warm sounding and you also lose resolution.
You have purchased an Aragon 8008BB, which is a very nice amplifier. It is biased high into Class A (the initial 26 watts if I remember right). Also, this was designed by D’Agostino and he likes the warm/rich FET type of sonic signature. I have heard this amp and it is very warm and rich sounding. However, as a byproduct of the warm signature with the Class A bias, it is definitely on the warm/slow side and you can sense a slight smearing of the sound. Nothing wrong with this -- it’s part of its sonic signature.
So, based on that, I would look for a very fast accurate bipolar-based preamp, such as the Rowland Capri you mentioned. The OPA1632 is a bipolar opamp with very good specs. It is going to be very clean and very fast. I would avoid Class A or tube preamps with your Aragon because there is going to be too much smoothing of the waveforms going through several Class A circuits and you will have a slower sound and the high frequencies may start to roll off. But if that’s what you want, then great.
I remember the older B&K stuff being very warm. It is actually too warm for me. You might want to look for a transparent/clean type of preamp. Cambridge Audio Azure 840E-B? Or the Wyred 4 Sound STP-SE?
The Freya uses a JFET solid-state stage which will add a warm/rich signature.
The Freya uses discrete J-fets as it's output stage, the opamps you see are most probably for dc servo duties because it's dc coupled, which even the $$$$$$K discrete preamps have.
From Schiit: "JFET Buffer Gain: 1 Frequency Response: 20Hz-20Khz, -0.2db, 3Hz-500KHz, -3dB THD: <0.001%, 20Hz-20KHz, at 2V RMS IMD: <0.0015%, CCIR SNR: >110db, A-weighted, referenced to 1V RMS Output Impedance: 75 ohms SE, 600 ohms balanced Topology: Single pair JFET per phase with passive distortion cancellation, DC coupled"
There are many different types . I have had many in my day .Currently I have which is considered the purest preamp type, Single ended . Which has several unique features . It is a Direct Heated Triode type completely Lundahl Transformer coupled 5 transformers,and 3 chokes and the little known 4P1L Vacuum tube which is very Linear, plenty of gain to eliminate the need of another tube for additional gain stage like many indirect Heated tubes. A very detailed and well balanced tube .it can be a bit microphonic in vibration if the builder does not isolate the tubes and isolate and damp the power from the Audio output section properly .modern technologies like relay remote and 63 step Shunt attenuated- Khozmo type. just another twist for another Audiophile dilemma.
You are right that opamps have become very good in this new millennium.
None of the aspects of circuit design/topology are mutually exclusive. There are some opamps that run class A natively. Any op-amp can be biased into class A Tube amps can run class B or Class A/B Almost any circuit can be designed to run with balanced I/O's or even better yet designed full balanced. It is also possible to design a circuit in one chassis yet make it run as a dual-mono unit.
Hi @paulburnett, several years ago, I had the original Rowland Capri in my system for about 15 months -- a truly excellent preamplifier, which was a significant upgrade from the ARC Ref3 that I used before. The original capri was audibly superior to Ref3 on most all parameters: staging and imaging, authority and macro/micro dynamics, resolution and extension, harmonic complexity, as well as sweetness and control of treble intermodulation. The only marginal flaw I could detect is a certain shyness of fundamental frequencies in the deep bass... Although Capri had tremendous authority, deep bass notes were a little shy of fundamental harmonics. But at less than 1/3 the list price of Ref3, I could not complain... By contrast, Ref3 bass sounded somewhat muddled.
The original Capri was indeed based on the OPA1632, and like all Rowland preamplifiers, runs balanced in class A, with transformer coupled inputs and outputs implemented with Lundahl transformers. I call Capri a preamplifier in stead of a pure linestage, because you can add to it a phono card.... and in more recent times a DAC card instead of the phono.
The newer Capri S2 is outwardly almost identical, except for a new display and remote control... However, the circuit has been redesigned, and the OPA1632 has been replaced. I have heard Capri S2 at length at RMAF, and it is sonically significantly enhanced over the original.... The first thing I noticed is that the deep bass shortcoming is totally gone.... Deep bass is, well... Deep, without ever becoming bloated.... And overall Capri S2 is even more musical than the original.
Guido Let's get real anyone who is running a tube PRE amp into a low input impedance amp should be alert that they won't achieve bliss. For you to claim op amp Pre amps out perform a ref tube unit without stating the conditions is misleading and wanted to point this out.
One more clarification.... Both Ref 3 and Capri were fully broken in.... REF 3 had a few thousand hours on it, and had fresh but broken in tubes for the SET Winged 6550 in the power supply and the 6H30s... Particularly important in the 6550... It takes this tube about 500 hours to reach full performance, and by 1000 total hours, it starts to sag.
Capri takes about 800 hours to reach maturity... Fresh out of the box, Capri is somewhat flat and blanched, and is easily outperformed by a well broken in Ref3.
After a quick initial comparison with Capri fresh out of the box, in which Ref3 was clearly superior, I did not apply Ref 3 again until completion of Capri break-in, so not to wear tubes down.
anyone who is running a tube PRE amp into a low input impedance amp should be alert that they won’t achieve bliss.
Just a FWIW, we make two balanced tube preamps and both can drive 10K no worries. They have a direct-coupled output so no loss of bass driving lower impedances.
For you to claim op amp Pre amps out perform a ref tube unit without stating the conditions is misleading and wanted to point this out.
I agree- I have yet to see an opamp based preamp that can keep up.
The tube route is always possible, but I like the idea of a no-fuss, very clean ss, valuing sound stage and detail over warmth,
If you like soundstaging abilities, so far I've yet to see a solid state preamp that can keep up with a good tube preamp in that regard. It always seems that low level detail suffers on the solid state preamps.
My apologies Johnny, your automotive metaphores are quite impenetrable.
Per the following page:
http://www.aragonav.com/8008-bb-mondial
The Aragon 8008BB has a balanced input impedance of 44KOhms, which is commensurate with the 40KOhms of three of my amps. It does not appear to have the 100KOhms input impedance that you asserted.
Interestingly, the input/output impedance ratio of my 40KOhms amps vs Ref 3 -- as well as versus ARC LS28 -- is
40,000Ohms / 600Ohms = 66.66 (periodic),
which exceeds the minimum safe threshold of 20 by a factor greater than 3. This strongly suggests that my Rowland amps are comfortably compatible with REF3. Similarly, the corresponding ratio on the Aragon 8008BB is
44,000Ohms / 600Ohms = 73.33 (periodic),
which is not staggeringly different.
The hybrid ARC LS28 is doubtless a lovely linestage. I have heard it at RMAF, and liked it. From my admittedly non-tuboholic point of view, with its SS power supply, it has the advantage over the REF series of limiting retubing to every couple of thousand hours or more for the 6H30s, as there is no fast-burner 6550 in the power supply which in old REF3 required a replacement every 1K hours or so.
It is worth pointing out that at $7500 new, LS28 is almost twice as expensive as the Rowland Capri S2 at $3995. From sheer memory, and not having heard the two linestages in the same system, it was not apparent to me if LS28 yielded any audible advantage over Capri S2. On the other hand, a side by side comparison of a well broken in pair in the same system might shed some light about any relative audible advantages of one over the other. However, preferability is hardly an absolute: personal musical taste and sonic priorities of the end user play a major role in every evaluation.
If you like soundstaging abilities, so far I've yet to see a solid state preamp that can keep up with a good tube preamp in that regard. It always seems that low level detail suffers on the solid state preamps.
I have noticed that my tube amp has a much deeper and wider sound stage than my AB amps. When I tried a Class D amp, the sound stage was incredibly shallow. Is this due to the damping factor (output impedance) or something else?
This is a great discussion, and I am grateful for all of the advice and opinions. Concerning impedance matching: I know about the 20x rule of thumb for input impedance of the amp over the output impedance of the preamp. If the dividend is in the ballpark, does the impedance ratio really matter to the sound? What does a low or high ratio roughly within the "20 times ballpark" sound like?
I shouldn't rule out tubes. I've never A/Bed good tubes vs. good solid state in my system, and I probably should. I just know I would be neurotic about swapping out tubes, worrying about when to do so. It's bad enough worrying about needles! ha ha
Actually that rule of thumb is 10:1. So if your output impedance is 500 ohms you should be able to drive 5000 ohms.
There is a complication with many tube preamps, which is the size of the coupling capacitor at the output. The issue is that to drive 10K ohms (which most of the time is the minimum you will see in a consumer amplifier) you need a pretty big coupling cap at the output of the preamp if you want it to play bass (which, if the cutoff is 20Hz, will be 10uf; but to avoid phase shift in the audio passband 100uf is recommended. You're not going to find very many preamps with a value that large- they usually split the difference in some way with 20uf being about as big as you're going to see). The bigger the cap, the more its going to color the sound so every manufacturer has a different hedge on this. For example, ARC noticed that many solid state amps have a higher input impedance and so limited their preamps to about 30K as a minimum in order to play bass right. Tube amps are generally 100K and some tube preamps have commensurately smaller coupling caps as a result, ignoring the solid state market altogether.
If there are tests available that plot the output impedance, you can see this by looking at the 20Hz output impedance as opposed to 1000Hz which is usually the frequency at which this spec is usually stated. The frequency plot will tell the story- often the 20Hz value will be multiples of the 1000Hz value. If its still 1/10th of the input impedance of the amp you're OK.
We avoided this by developing a direct-coupled output circuit for which we got a patent. As a result the output impedance vs frequency curve is a flat line- the same at 10 or 20Hz as it is at 1000Hz or 10KHz.
Just wanted to mention a few of you have given misinfo on the schiit freya, it doesn’t only have a Jfet solid state output stage, it actually has 3 seperate output stages you can change between. Fully passive, jfet buffered 0 gain, and tube gain output
Specifications may or may not (most often not) relate to great sound quality. 2 different preamplifiers that measure similarly will often sound very different. If you limit your search to "big name brands" you may be missing some opportunities. To say nothing of the fact that as your original question regarding "design philosophies & circuit topology" render a cloudy picture. As a sidebar the audio industry has no standards on anything. Even preamplifier output values. IMO preamplifiers are the weakest link in the audio chain. I have always been a proponent of full-featured preamplifiers and single ended vs balanced. Unless you have no aspiration on spinning vinyl then you can simplify your search to line level devices. Concentrate on in home auditions and listen. Good luck!
I would just cut to the chase and go Pass Labs. The integrity and dedication of this man over the years, and his willingness to share with the DIY community, and the consistent quality of his product makes it almost a no brainer IMO. I have a class A amp of his and everything about it is just simply art in every respect-- which of course its sound reproduction is paramount. I have cut back on my "habit" as I am working on making a major move out of the area otherwise I am itching for a Pass preamp.
^ I agree on Pass. My first the Threshold Stasis 3, then the Aleph 3 and 5, then the great X150 (probably the best used 'cheapish' general amp one could locate).
The P and L and Xono and others have been great preamps, LS's, and phono stages as well. I've 'slummed' for 20+ years with the Pass designed GFP-750, which died (accidental).
Retired now, so the budget has taken a hit, so the Freya looks like a good choice to go w/ my X-150, Schiit Gumby. Been using the Ragnarok as a line stage, so the Freya will be a step up.
In terms of Schiit, like to see them make an amp that runs class A up to 25 watts, and A/B up to 100+. The Vidar (in mono) and Ragnarok do not handle loads under 4 ohms.
In terms of tubes. I like them in pre-amps (ARC SP-15 onward) but never could stomach tube amps - even very expensive ones in the bass, and retubing cost? My lord...
you can always spend a bit fixing up the pro 10mc, refreshing its capacitors with newer higher spec ones (minimally...all the old electrolytics have to go)..and then look at changing the opamps in it to higher spec newer designs.
this is the least expensive ticket with the highest return, in your current scenario.
Manufactures using op amps makes all the sense in the world to me. We all preach that the shorter the circuit path, the less involved, the better sounding....well really anything. As these micro parts improve, incredibly short signal paths of opamps actually should sound better than the same circuit wired by hand. So, If a smart circuit design can be shoved into an op amp??? Someday, op amps will be the norm in many high end pieces of equipment.
the problem for me, is that they tend to be very ’mono’ in their dynamic expression. Meaning the dynamic expression is very one note or one singular expression of dynamics.
Pinched, is another word to describe their sound qualities.
Where real music has a plethora of dynamic responses and behaviors that all move in and out of one another..
It is difficult if not impossible to get past this as the traces are fine and everything is made out of silicon, so all the bits of the sonic presentation - sound likened to that.
It takes a bit of time to hear it, but it can’t be unheard, once realized.
This is where discrete components come in and people buy discrete circuit audio gear.
Opamps used as well as they can be, does indeed move up the ladder of sound quality and can and do sound quite good. But if we are talking about pinnacles, and trying to get to pinnacles, they don’t apply, in my experience or expectations.
There is a lot of overlap in these qualities of opamp vs discrete circuits, thus arguments and discussions are born. A not so perfect discrete circuit can be eclipsed by a well implemented opamp circuit.
When all things are equal, which they never are..in that case, discrete is better than opamps.
Twist! I just found out that the Schiit Freya is only balanced out from the tube stage, and further, that my Aragon 8008BB shorts the balanced input when using single-ended (jumpers in place on the balanced ins if using RCA), which means I would really need to love only the tube stage and never use either the passive or jfet buffer. It's a bit of a baffling design choice. I guess Mike Moffat really love his tubes. And yet, many tube-only preamps are also single-ended only.
Will Schiit ever consider making a fancier pre that's balanced all the way? Throw the Mani's guts into a shielded corner? I'd pay a fair bit more for that.
@paulburnett "...the Schiit Freya is only balanced out from the tube stage...."
Paul - I'm not sure that accurately represents Freya functionality. There might be some confusion based on a statement in the Freya manual. See the Freya connections page and this note under "2. Single-Ended Inputs"...
Note: Freya
does not convert single-ended to balanced,
except when using the
tube gain stage.
My understanding is that this statement applies to single ended inputs only. If you are using a singled ended connection from source to Freya with only a balanced connection from Freya to a power amp, you won't get sound except in tube mode.
I have asked Schiit for confirmation.
In my experience, balanced sources connected via XLR work fine using the Freya's balanced out connection whether running Freya in Passive, JFET or tube mode.
I agree with your interpretation. I have a balanced amp that accepts either balanced or RCA, but not both at the same time. I listen to vinyl 50/50 with digital, the former of which comes in single-ended. I was looking forward to being free to listen in tube, jfet, or passive, but if I understand correctly that's not an option if I run balanced to the amp, which is what I want to do. If Schiit has options other than running RCA to the amp or listening through tube stage only, I will be excited to hear what you learn.
Own a Freya myself and might have a similar set up. Phono pre for vinyl is singled ended; digital (DAC & CDP) is balanced - so there are singled ended and balanced inputs to the Freya. The Taranis power amp is balanced only. Freya connects to it via an XLR IC. I switched out the Taranis (just for a change of pace) a little while ago but I was definitely able to listen to digital (Bal. in) with Freya in JFET mode and and switch to vinyl (SE in) using Freya in tube mode all while being connected via XLR only from Freya to the balanced input of the Taranis.
I do NOT recall whether I got sound from the SE/vinyl input if I tried to listen in anything but tube mode. Not sure I tried. Apologies if I'm being a little dense and missing your point. I don't mind re-installing the Taranis to confirm things if this will address your concern. Let me know.
Y'all know that a phono cartridge is a balanced source right?
Why all this concern about balanced operation between the preamp and amp, when the more important place (if you plan to listen to phono) is the phono cable?
The balanced line system was created to eliminate interconnect cable artifact. I can't think of a better place to do it than at the point where the signal enters the system.
This preamp has everything you say you are looking for in a sonic signature and more. Only caveat is that the dealer ad does not say whether or not it has the optional phono stage. I found it to be worthwhile to build a XLR-terminated wiring harness from cartridge to the phono stage's XLR inputs as atma-sphere says.
No affiliation with seller but I have owned this preamp (with the optional phono stage) for 12 years and it is a keeper IMO.
You must have a verified phone number and physical address in order to post in the Audiogon Forums. Please return to Audiogon.com and complete this step. If you have any questions please contact Support.
